Okeechobee, Florida. 1969-72”, U S . Geological Survey Report Investigation No. 71, Tallahassee, Fla., 1974. (14) Mattraw, H. C., Jr., Sherwood, C. B., J . Res. U.S. Geol. Suru., 5,823 (1977). (15) Burton, T. M., Turner, R. R., Harriss, R. C., Florida State Universitv. Tallahassee. unDublished data for 1974-1975: Dersonal communTcation, 1978. (16) Bourne. R. G , M.S. Thesis, University of Florida, Gainesville, 1976. (17) Odum, H. T., Ewel, K. C., Ordway, J. W., Johnston, M. K., Third Annual Report to the National Science Foundation and the Rockefeller Foundation. Center for Wetlands. Universitv of Florida, Gainesville, 1976. (18) American Public Health Association. “Standard Methods for the Examination of Water and Wastewater”, 14th ed., Washington, D.C., 1975. (19) Environmental Protection Agency, “Methods for Chemical Analysis of Water and Wastes”, National Environmental Research Center, Cincinnati, Ohio, EPA-625/6-74-003a, 1976. (20) Junge, C. E., “Air Chemistry and Radioactivity”, Academic Press, New York, 1963. (21) Brezonik, P. L., Shannon, E. E., “Trophic State of Lakes in North and Cenmal Florida”, Publication 13, Water Resources Research Center, University of Florida, Gainesville, 1971. (22) Vollenweider, R. A., “Scientific Fundamentals of the Eutrophication of Lakes and Flowing Waters, with Particular Reference t o Nitrogen and Phosphorus as Factors in Eutrophication”, O.E.C.D. Report, DAS/CSI 68.27, Paris, 1968. (23) Vollenweider, R. A., Schweiz. Z Hydrol., 37,53 (1975). (24) Brezonik, P. L., in “Nutrients in Natural Waters”, Allen, H. E., Kramer, J. P., Eds., Wiley-Interscience, New York, 1972.
(25) Wolaver, T. G., “The Distribution of Natural and Anthropogenic Elements and Compounds in Precipitation across the US.: Theory and Quantitative Models”, EPA, Division of Ecological Research, Research Triangle Park, N.C., Technical Report, 1972. (26) See Cooper, H. B. H., Demo, J. M., Lopez, H. A., in ref 1. (27) Hendry, C. D., M.S. Thesis, University of Florida, 1977. (28) Dawson, G. A., Atmos. Enuiron., 12,1991 (1978). (29) . . Chameides. W. L.. Stedman. D. H.. Dickerson. R. R.. Rusch. D. W., Cicerone, R. J:, Atmos. Sci., 34; 143 (1977)’. (30) Granat, L., Tellus, 24,550 (1972). (31) National Research Council. “Air Qualitv and Stationarv Source Emission Control”, Commission on Natural Resources, National Academy of Sciences, Ser. No. 94-4, U S . Government Printing Office, Washington, D.C., 1975. (32) Liljestrand, H. M., Morgan, J. J., Enuiron. Sci. Technol., 12, 1271 (1978). (33) Gorham, E., Q. J . Roy. Meteorol. Soc., 84,274 (1958). (34) Barr, A. J., Goodnight, J. H., Sall, J. P., Helwig, J. T., “A Users Guide to Statistical Analysis System”, SAS Institute, Inc., Raleigh, N.C., 1976. (35) Frohliger, J. O., Kane, R., Science, 189,455 (1975). (36) Galloway, J. M., Likens, G. E., Edgerton, E. S., Science, 194, 772 (1976).
Receiued for review July 19, 1979. Accepted March 26, 1980. This work was part of a project of H. T. Odum (principal inuestigator),, the University of Florida Center for Wetlands,jointly sponsored by the R A N N Division of the National Science Foundation and the Rockefeller Foundation.
Respirable Aerosols from Fluidized Bed Coal Combustion. 1. Sampling Methodology for an 18-Inch Experimental Fluidized Bed Coal Combustor George J. Newton”, Robert L. Carpenter, Hsu-Chi Yeh, and E. R. Peele Inhalation Toxicology Research Institute, Lovelace Biomedical and Environmental Research Institute, P.O. Box 5890, Albuquerque, N. hAex. 87115
Fluidized bed combustion of coal, lignite, or other materials has a potential for widespread use in central electric generating stations in the near future. This technology may allow widespread use of low-grade and/or high-sulfur fuels due to its high efficiency a t low combustion temperature and its ability to meet emission criteria by using limestone bed material. Particulate and gaseous products resulting from fuel combustion and ielutriation of bed material are discharged through an exhaust cleanup system. Equipment used to obtain aerosol samples from the exhaust system of the 18-in. fluidized bed combustor (FBC) at the Morgantown Energy Technology Center (METC) is described. Selection of sampling sites led to design of an aerosol sampling train which allowed a known quantity of the effluent streams to be sampled. A 15 to 25 L/min sample was extracted from the duct, immediately diluted, and transferred to a sampling/aging chamber. Transmission and scanning electron microscope samples, two types of cascade impactor samples, vapor-phase and particulatephase organic samlples, spiral duct aerosol centrifuge samples, and filter sample!; were obtained. The sampling system was designed to have (anupper size limit of about 10 pm aerodynamic diameter. Samples have been obtained from the FBC while Montana Rosebud subbituminous coal, Texas lignite, Western Kentucky high-sulfur bituminous coal, and Paraho oil shale were burned. Methods and equipment used to characterize both particulate and gaseous phases of aerosols in flue gases from an experimental 18-in. (i.d.) atmospheric pressure fluidized bed coal combustor are the subject of this report, the first paper in a 0013-936X/80/0914-0849$01 .OO/O
series of reports on physical and chemical assessment of airborne effluents from fluidized bed coal combustion. Our reports are primarily concerned with trace constituents of these airborne effluents. Major constituents of flue gases and fly ash from both conventional pulverized coal and fluidized bed coal combustion have been widely reported. Analyses are further restricted to respirable constituents of these exhaust aerosols. Combustion of coal, anthracite waste (culm), lignite, or other materials in a fluidized bed combustor (FBC) has unique advantages over conventional combustion technologies. Conventional pulverized coal combustion involves mixing finely ground coal with air and injecting the air/coal mixture into the combustion region through a nozzle. Finely divided coal burns with nearly explosive rates, and flame temperatures may approach 2500 OF (1400 O C ) . At these temperatures, oxides of nitrogen are produced in high quantities, and sulfur, if present, is oxidized to sulfur dioxide. Both nitrogen oxides and sulfur dioxide are major pollutant problems. No proven economic technology is available for NO, control, and SO2 removal from the effluent stream is costly. Fluidized bed combustion using wider ranges of fuel types can meet present EPA emission standards for both SO2 and NO,. The fluidized bed, with lower operating temperatures (800-900 “C), produces much lower levels of nitrogen oxides. Since the presence of noncombustible materials in the fluidized bed is essential, material such as limestone may be added to adsorb Son. Other advantages of FBC are enhanced heat transfer due to immersion of heat exchangers in the fluidizing bed, lower construction costs, and potential to use fuels with heat content as low as 2900 Btu/lb (6.72,X lo6 J/kg).
@ 1980 American Chemical Society
Volume 14, Number 7 , July 1980 849
FLUID- BED
COMBUSTOR
r
Ash Bed Moterial-+
--
Feeder
Flgure 1. Schematic of fluid-bed combustor showing fuel and limestone
feed systems and the three cleanup devices Beyond determination of SO, and NO, emission levels, little is known about the potential human health risks associated with fluidized bed coal combustion. In order to address these questions, a cooperative biological assessment effort between the Lovelace Inhalation Toxicology Research Institute (ITRI) and the Morgantown Energy Technology Center (METC) was initiated. Goals of the sampling effort were to determine: (1)respirable particle mass loading, (2) size distribution parameters, (3) trace elemental characteristics of particles, (4) organic constituents present in the vapor phase or associated with particles, and ( 5 ) specific surface area based on adsorption of nitrogen. These parameters were determined at various locations from the top of the combustor to the stack breech. Description of M E T C Fluidized B e d Combustor
The 18-in. (i.d.) atmospheric pressure fluidized bed combustor has been under development at the Morgantown Energy Technology Center since 1967. Various fuels including coals, lignite, and culm have been burned in the combustor. Bed materials have also been varied in efforts to enhance SO2 capture. Figure 1 is a schematic of the 18-in. fluidized bed combustor. The fluidized bed is supported by a conical plate perforated by 0.125-in. holes fitted with stainless steel elbows. These elbows direct fluidizing air toward the center of the vessel parallel to the conical surface. Water passing through the heat exchanger bank, which consists of U-shaped tubes (0.25-in. i.d.), extracts heat from the bed. An additional heat exchanger at the top of the combustor vessel controls the temperature of exiting gases. A screw conveyor meters presized coal into a pneumatic feed tube, which conveys fuel into the combustor near the bottom of the bed. Limestone or other sorbent may be fed into the combustor by a screw conveyor above the fluidized bed. Bed level is controlled by periodic removal of material from the bottom with a 3-in. screw conveyor via a lock hopper. Combustion products pass through two cyclone separators. Solids from the primary cyclone contain enough unburned fuel to justify reinjection into the combustor. Partially cleaned combustion gases pass through a baghouse filter system for final particle removal. A system of valves, illustrated in Figure 1,allows the baghouse to be placed in the cleanup system or to be bypassed. D a t a Uses
The purposes of the data base are: (a) to characterize airborne constituents in the effluent stream as a basis for eval850
uating control technology and making potential changes in system designs to improve control and (b) to characterize respirable constituents for use in evaluating potential health risks to workers from fugitive emissions or to the general population from environmental release. The FBC was sampled at four positions (1-4 in Figure 1). Samples obtained at position 1 (stack breech) were of greatest interest to those concerned with atmospheric releases. Samples from positions 2,3, and 4 were representative of source material for potential fugitive emissions to the work place environment. Samples from all four positions provide an engineering assessment of the combustion process and the effectiveness of cleanup devices. Major emphasis was placed on aerodynamic size selective sampling because of the critical importance of aerodynamic size in predicting lung deposition of inhaled particles. Similarly, chemical analysis of aerodynamically sized samples was a major goal. Preliminary results of sampling and chemical analysis have been described ( I ) .
Environmental Science & Technology
Materials a n d M e t h o d s
Basic Considerations. A major limitation in understanding coal combustion effluents is an inability to adequately sample the aerosols. Since no single aerosol instrument currently in existence is capable of meeting the stated goals, a series of instruments and techniques was used. Flow conditions within ducts were very turbulent with flow Reynolds numbers on the order of lo4-lo5. These turbulent flow conditions have no defined streamlines. Since the concept of ‘“isokinetic sampling” is designed to minimize sample bias at the probe inlet by minimizing streamline perturbations, isokinetic conditions cannot be defined for turbulent flow. With turbulent flow conditions, matching mean probe velocity to duct velocity will not be “isokinetic” and probably should be called isomean-velocity (IMV) sampling. As an accommodation, probe sample velocities were selected so that the ratio of duct to sample velocities ranged from one to two. In addition to turbulent flow, preliminary reports indicated that temperatures could range from 800 “C at position 4 to 150 “C at position 1. Although sampling probes were designed for these temperatures, actual temperatures encountered were 600 to 150 “C at the respective positions. Two approaches were considered: (a) sample at stream temperature within ducts or (b) extract a sample, dilute, age, and sample under milder conditions. Sampling at temperature within ducts was rejected due to small duct size, material corrosion and erosion, potential “cloud effects” on size-selective samplers from high mass loadings, and high temperature effects on aerosol parameters. Specifically, the effect of temperature on the “Cunningham” slip correction factor is unknown, although several estimates of this effect have been made (2,3).Basic strategy involved extracting a small sample (15-25 L/min) with duct to sample velocity ratios of about 1.2:1, diluting with 100 L/min of dry cool air, and piping the diluted sample to a chamber from which samples could be obtained under milder flow conditions. Sampling System Design. Figure 2 is a schematic diagram of major sampling train components. Sampling probes were 3-ft lengths of 0.5-in. 0.d. (0.062-in. wall) stainless steel tubing with inserts at the probe tip to reduce the inside bores to 0.375 in. (0.95 cm) for positions 1 and 2 and 0.25 in. (0.64 cm) at positions 3 and 4.Edges were chamfered to reduce edge effects, At positions 1 and 2 a 90” bend was used to align the probe in the 8-in. i.d. stack, while a t positions 3 and 4 the probes were installed on the outer radius of an elbow SO that about 24 in. (60 cm) of the sample probe was axially located inside a straight section of the 4-in. (10 cm) i.d. duct. A ball valve was placed immediately after the sample port, permitting the sampling system to be detached while maintaining FBC exhaust integrity.
SAMPLING INSTRUMENTS
omple
Flow Exhourl
Diluting Air
Flow Conlrol
I
Sornplr
Flow SAMPLING CHAMBER
rl
LT-T~ EXHAUST FLOW
lFlller I
Sample-
CONTROL
Flow
?-I
Compressor
I
I
Figure 2. Schematic diagram of sampling system major components
HaDIALLY INJECTED DILUTER Mal'l: 302 S.S.
k~''NPT
I i I..'. Gasket /
u
B 0 Q
'4 repa- 90.Apart
t in
Y
Air
Figure 3. Cross-sectional view of radially injected diluter used to reduce thermophoretic losses
Dilution of hot exhaust gases generates thermal gradients which could cause fine particle losses in dilution systems. Since fine particles move toward cooler surfaces when subjected to a temperature gradient ( 4 ) ,these thermophoretic losses were of concern. Therefore, a special radially injected diluter (Figure 3) was designed to reduce thermophoretic losses. Aerosol enters the diluter through a 0.5-in. (1.27 cm) pipe. A high-temperature gasket separates the dilution region from the support flange. Diluting air enters through a porous stainless steel cylinder perpendicular to the sample flow direction. This arrangement tends to compress the sample, allowing cooling and mixing to occur away from the cooler walls and thereby reducing thermophoretic losses. Examination of the diluter after several days of operation demonstrated that losses were minimal. As shown in Figure 4, four compressed air lines were connected to the diluter assembly, each containing its own flow controls. Final filters were used to ensure that no particulate matter was introduced into the sample by diluting air. After exiting the diluter, the sampling train was expanded to 1.5-in. i.d. copper pipe for transfer to the sampling chamber which was a specially adapted 55-gal steel drum with appropriate fittings. Transport line diameters were chosen so that flow Reynolds numbers were in the transition region (2000-4000) to minimize losses ( 5 ) . Immediately downstream from the diluter, a Tenax-GC (Altech Associates, Arlington, Ill.) sampling system was installed in a tee of the sampling line. Figure 4 illustrates this Tenax sampling system as well as the dilution airflow controls. Tenax, a gas chroinatography column support material, has the property of adsorbing gaseous hydrocarbons larger than the Cq isomers. Since its initial use by Zlatkis et al. (6), this material has been increasingly used as a probe for gaseous
I
&Filter
Lotametars
'Compressed Air
Figure 4. Schematic diagram of dilution air flow controls with diluter and Tenax samplers Lovelacs Multijet (Impactor
Smal I Sampling Tee fw Low Flow
P 1
To V a c y
'2O*kompling
Devices
6Mapnehelic
Sampling Chamber asKr Source
Figure 5. Schematic diagram of sampling chamber showing internal
structure and sampler flow controls hydrocarbons. The Tenax side of the sampling line contained a cutoff valve to maintain the integrity of the sampling system, a filter to eliminate particulate material, and a brass cylinder containing the Tenax adsorbent. Exhaust from the Tenax adsorbent line was passed through the sample cleanup filter before exhausting into the main vacuum line. The sampling chamber (Figure 5 ) served several functions. The sample entry line was constructed so that particles larger than 10 pm aerodynamic diameter would be removed by impaction on the sampling chamber walls. The sampling chamber was fitted with ports permitting sampling instruments to be connected with minimum flow disturbances. Excess sample was pulled through the chamber and then Volume
14, Number 7, July 1980
851
passed through a concentric electrostatic precipitator and an absolute filter before being vented outside. Figure 5 also details the structure of the sampling chamber. The sampling line coming from the probe entered vertically through the top of the chamber and was turned perpendicularly so that the gas stream is aimed at one wall. Provisions were made for both low-flow (0.1-1.0 L/min) and high-flow (25 L/min) sampling instruments. Sample lines extended through the top of the barrel and part way down into the chamber. The position of these lines was chosen so that the incoming sample was pulled up past these lines by the chamber exhaust line. The sampling chamber contained an 85Kr deionizer source to reduce electrostatic charge on particles to near Boltzmann equilibrium (7, 8). A blowout port was present for safety purposes, and two sampling chamber pressure gauges were included. One gauge measured sampling chamber pressure with respect to the sampling probe, while the other measured sampling chamber pressure with respect to atmospheric pressure. These two gauges were necessary to ensure transport of the sample from the effluent line to the sampling chamber. Pressure gauges also ensured that the entire system operated at desired pressure relative to atmospheric pressure. Sample and exhaust flows and pressure were controlled by using rotameters and pressure gauges. A portable air compressor system with associated air cleanup and drying units was used as a source of diluting air. Two rotary vacuum pumps were used to exhaust the sampling chamber. The sampling chamber was maintained at a slightly negative pressure and the amount of effluent drawn from the stack was controlled by proper adjustment of dilution and exhaust flow controls. The airflow control system had three distinct parts: (a) a diluter flow section (Figure 4), whose function was to ensure that an accurately metered flow of air was supplied to the diluter, (b) the main exhaust control (Figure 5 ) , which consisted of a group of pressure gauges and flow controllers in parallel as well as associated cutoff valves to accurately measure flow leaving the sampling chamber, and (c) sampling instrument controllers (Figure 51, a series of parallel flow controllers and pressure gauges. The sampling instruments withdrew varying amounts of air from the sampling chamber. In order to maintain a constant flow out of the sampling chamber it was necessary to set up the exhaust flow controllers such that several of them could be preset to control the same flow that each of the sampling instruments used. Each sampling instrument has its own flow controller, which was preset to the appropriate flow. The valve system permitted one to simultaneously begin sampling with an instrument and cut off the appropriate flow from the main exhaust control section. Diluter airflow controls and chamber exhaust controls were set so that 100 L/min of air was injected into the diluter, while 125 L/min was withdrawn from the sample barrel. The difference, 25 L/min, represented the probe sampling rate. Flow controls were adjusted so that the sampling chamber was negative with respect to the probe and negative relative to the atmosphere. To sample the effluent aerosol, the appropriate sampling instrument was installed in a sampling tee and connected to the designated sampling flow controller, and the appropriate fraction of flow was switched from the chamber main exhaust to the sampling system, initiating flow through the sampler. Sampling System Collection Characteristics. Conventional sampling probes usually have an L-shaped configuration where a particle laden gas stream entering a probes makes a 90" change in direction within a short distance and is then transported to a collecting or detecting instrument. When gas stream velocity is very high and/or flow is turbulent, isomean-velocity (IMV) sampling will reduce sample probe bias 852
Environmental Science 8. Technology
0.8 /so meon-velocity Samp/ing 1
$ 0.4 v
o.2
t
I
I
1.0
5
10
AERODYNAMIC DIAMETER, p m
Figure 6. Predicted sampling probe efficiency (inertial effects only) for aniso-mean-velocity and iso-mean-velocity sampling conditions (Lshaped probes) on the experimental 18-in. FBC. Conditions: temperature = 150 "C, total flow rate = 500 scfm, duct diameter = 20.32 cm, sampling flow rate = 25 slpm, probe diameter = 0.95 cm (AIMV) and 0.85 cm (IMV)
as compared to aniso-mean-velocity (AIMV) conditions, but bias due to particle losses at the bend may be so high that the overall bias is worse than encountered in a well-designed AIMV sampling system. Figure 6 shows an example of predicted penetration through one of our probes operated under AIMV conditions and a similar analysis of so called "isokinetic" or IMV conditions. Predicted aerosol penetration was based on a method of Davies (9) for calculating sample entry efficiency and Yeh's (10) equation for calculating aerosol losses at a 90" bend. In addition to sampling efficiency, several factors were also considered in the design of our sampling system. These considerations include: (a) adaptability for sampling under a variety of operating conditions of the experimental FBC, (b) probe installation within existing exhaust ducts, and (c) offthe-shelf availability of probe components. Table I lists probe sizes used, sampling conditions, and an estimated sampling efficiency using Davies' equation for aniso-mean-velocity and Yeh's equation for aerosol losses at a 90" bend. A system similar to that used on the FBC was assembled in the laboratory. Pipe sizes and horizontal distances were duplicated, as was the sampling chamber configuration. A compressed air nebulizer generated a test aerosol from a suspension containing 30 mg/mL montmorillonite clay and 10 mg/mL dissolved sodium fluorescein dye. The resultant aerosol was characterized as approximately log-normal with a mass median aerodynamic diameter (MMAD) of 2.5 pm and a geometric standard deviation (cg) of 1.8. The sampling system was operated for time periods comparable to those used for FBC sampling. For probes with a 90" bend, maximum total system losses did not exceed 17%. For straight probes total losses never exceeded 12%. Measured losses due to the 90% bend were about 570,with the remaining 12% diffusely distributed throughout the sampling lines.
Results and Discussion The sampling system described herein was placed in operation in early 1977. Since that time all sample types listed in Table I1 have been obtained without significant problems. Four fuel types have been studied: (a) a subbituminous coal, Montana Rosebud, (b) Texas lignite, (c) Western Kentucky high-sulfur bituminous coal, and (d) Paraho oil shale. These fuels were studied under a variety of FBC operating conditions and, in the case of Texas lignite, several fuel treatments were also studied. A modified version of the sampling system has also been used to sample process streams from METC's stirred bed Lurgi low-Btu gasifier ( 1 1 ) . Based on our experiences in respirable aerosol sampling
Table 1. Probe Sizes, Range of Sampling Conditions, and Estimated Probe Sampling Efficiency ( C/Co)for Use on the Experimental FBC duct size, 20.32 cm total flow 200 rate, scfm stream temp, 150
20.32
10.16
10.16
500
200
500
150
810
810
OC Re, duct probe size, cm sampling flow rate, slpm Re (probe) U (probe), cm/s Uo (duct)/U (probe)
2.82X lo4 7.04 x 104 5.63 x 104 1.41x 105 0.635 0 . 9 5 ~ ~ 0.635 0.951~ 25
25
2.65X lo3 2.65x 103 2.13 x 103 2.13 x 103 4551 793 4551 793 1.25
0.50
co
Dae, Ccm
1 .o 1 .o 0.99 0.94 0.86 0.74 0.52
0.1 0.5 1 .o 3.0 5.0 7.0 10.0 a
25
25
L-shaped probe.
Straight
0.89
c/co
c/ co
1 .o 1 .o 1 .o 0.97 0.92 0.84 0.65
1 .o 1 .o 1 .o 0.98 0.96 0.94 0.92
2.21 c/ co 1 .o 1.01 1.04 1.20 1.41 1.60 1.80
probe.
Table II. Samplers Used on Effluent Streams from METC FBC sampling device
Lovelace multi-jet cascade impactor (LMJ) Sierra radial slit jet cascade impactor (SRSJ) Mercer cascade impactor (MI) Lovelace aerosol particle separator (LAPS) filter (47and 90 mm diameter) point-to-plane electrostatic precipitator Tenax sampler concentric electrost,atic precipitator
flow rate, Lfmin
25 21 0.5
0.3 25 0.5 2.0 100-1 25
from several fossil fuel conversion plants and from mixedoxide nuclear fuel fabrication plants, several generalizations can be made. Sampling of process and exhaust streams via extractive techniques followed by dilution has proven to be a useful method for stabilizing the sampled aerosol. Coagulation, thermophoretic los,ses,and vapor condensation are minimized after extraction by this technique. Ambiguities in aerosol size measurements are reduced by dilution because temperatures are reduced and stabilized. Additionally, gas density and viscosity approximate those of air, eliminating the need to measure these variables. Unless procee,s conditions require laminar flow, process and exhaust streams will be highly turbulent. This results from efforts to maximize mass transfer while minimizing duct
or pipe sizes consistent with good engineering practices. Conditions usually encountered in full-scale plants result in calculated Reynolds numbers in the range of 10;to lo6. Under “isokinetic” or IMV sampling conditions, probe entrance and duct velocities are matched. Reasonable sampling flow rates to achieve IMV conditions frequently require small diameter probes which plug readily. High sample velocities in such probes result in excessive probe losses at bends, offsetting efforts to reduce sample bias at the probe inlet through isokinetic or IMV sampling. For straight probes, IMV sampling should minimize probe losses. However, losses in curved probes should be carefully evaluated and AIMV sampling may result in minimum overall losses. The series-train approach to sampling requires that all samplers operate at the same flow rate. Although series trains are more compact, the high degree of mechanical integration impedes on-line sampler substitution. Although less compact, the parallel sampler system described herein permits ready exchange of samplers during operation. Combined sampler flow rates may range from a few milliliters per minute to the total system flow (125 L/min). Any combination of samples described in Table I1 may be obtained simultaneously within these flow limitations. The mix of sample types is easily adjusted. Thus, during a single sampling effort, different aspects of a toxicological assessment can be emphasized. Acknowledgment
The authors thank METC personnel who supplied FBC design and operating conditions, especially Drs. J. J. Kovach for coordination of efforts and W. M. Wallace, G. D. Case, L. C. Headly, J. Wilson, R. Rice, and U. Grimm for constructive criticism and review. We also acknowledge the efforts of the ITRI sampling team, R. Tamura, D. Horinek, R. Peele, T. Stephens, and E. Barr. Finally, we thank our scientific colleagues at ITRI for encouragement, review, and support, especially Drs. C. H. Hobbs and R. 0. McClellan. L i t e r a t u r e Cited (1) Carpenter, R. L., Weissman, S. Heisler, Newton, G. J., Hanson, R. L., Peele, E. R., Mazza, M. H., Kovach, J. J., Green, D. A., Grimm, U., “Characterization of Aerosols Produced by an Experimental Fluidized Bed Coal Combustor Operated with Sub-Bituminous Coal”, Lovelace Inhalation Toxicology Research Institute Unclassified Report, LF-57, 1978. (2) Raabe, 0. G., J . Air Pollut. Control Assoc., 26,856 (1976). (3) Willeke, K., J . Aerosol Sci., 7, 381 (1976). (4) Waldmann, L., Schmitt, K. H., in “Aerosol Science”, Davis, C. N., Ed., Academic Press, New York, 1966, p 137. (5) Strom, L., Atmos. Environ., 6, 133 (1972). (6) Zlatkis, A., Bertsch, W., Lichtenstein, H. A., Tishbee, A,, Shunbo, F., Liebich, H. M., Coscia, A. M., Fleischer, N., Anal. Chem., 45, 763 (1973). (7) Liu, B. Y. H., Pui, D. Y. H., J . Aerosol Sci., 5,465 (1974). (8) Teague, S.V., Yeh, H. C., Newton, G. J., Health Phys., 35,392 (1978). (9) Davies, C. N., Br. J. Appl. Phys. Ser. 2, 1,921 (1968). (10) Yeh, H. C., Bull. Math. B i d , 36. 105 (1974). (11) Peele, E. R., Carpenter,R. L., Newton, G. J., Sturm, M. M.,Inhalation Toxicology Research Institute Annual Report, 1976-1977, LF-58. Inhalation Toxicolom Research Institute. Lovelace Biomedical and EnvironmentarResearch Institute, Albuquerque, N. Mex., 1977, p p 243-50.
Received for reuieu! February 2, 1979. Accepted March 31, 1980. Research performed under U S . Department of Energy Contract E Y-76-C-04-1013,
Volume 14,Number 7,July 1980 853